Abstract: Lightning location requires high real-time and positioning accuracy, and the grid search algorithm needs a large amount of computation. An optimization method of lightning location based on grid search is proposed. This method quickly searches the candidate target regions by using multi-dimensional spatial data index. Afterwards, comparing the classification error of SVM（Support Vector Machine） classifier which is stored on the target candidate regions for lightning location. Experimental results show that this method has well performance on positioning accuracy and reliability, and can meet the requirements of real-time lightning location.

Key words: lightning location, Support Vector Machine（SVM）, grid search, real-time

摘要： 闪电定位具有较高的实时性和定位精度要求，网格搜索算法存在计算量大的问题。提出了一种基于网格搜索的闪电定位算法优化方法。该方法采用多维空间数据索引快速搜索候选目标区域，通过对比存储在各候选目标区域上的SVM（Support Vector Machine）分类器的分类误差进行闪电定位。实验表明，该方法具有较好的定位精度和可靠性，且能满足实时闪电定位要求。

关键词: 闪电定位, 支持向量机, 网格搜索, 实时性